Week 2 progress - Jelc-sys/AV-control-handover-in-critical-situation GitHub Wiki

This week we made progress on our goals set during the meeting. The main aspects of our work were:

• Research on the implementation of waypoints for the AV.
• Research on different cues to alert the driver to take over (visual, auditory, haptic).
• Implementation of two visual cues (light flash and alert on dash) and one auditory cue (horn for now, but can be easily changed).
• Implementation of reaction time detection. The timer will start with the cue, and stop when the driver takes action. This was added with the assistance of Generative AI.

In addition to the implementation above, we spent a large amount of time learning about Unity and how to implement some aspects. Implementation is taking us longer than initially anticipated due to the difficulty encountered to learn a new software, combined with understanding the implementation of the coupled-sim. We hope that as we familiarise ourselves more with Unity, implementation will become faster and we would be able to make more progress per week.

The progress this week brings us closer to achieving the learning objectives of the course, as we are beginning to analyse the selected context. While, until now, the application was limited to the vehicle, we will begin working in the coming week on external factors, including pedestrians and other road vehicles, which could include emergency vehicles. We will also look into platooning as a potential implementation of AVs in the real world, and how a platoon will react to the driver taking over control of the vehicle. Lastly, we intend to begin programming the takeover procedure, which would involve the car alerting the driver to take over control in the critical situation described.